Deep artificial neural networks apply principles of the brain’s information processing that led to breakthroughs in machine learning spanning many problem domains. Neuromorphic computing aims to take this a step further to chips more directly inspired by the form and function of biological neural circuits, so they can process new knowledge, adapt, behave, and learn in real time at low power levels. Despite several decades of research, until recently, very few published results have shown that today’s neuromorphic chips can demonstrate quantitative computational value. This is now changing with the advent of Intel’s Loihi, a neuromorphic research processor designed to support a broad range of spiking neural networks with sufficient scale, performance, and features to deliver competitive results compared to state-of-the-art contemporary computing architectures. This survey reviews results that are obtained to date with Loihi across the major algorithmic domains under study, including deep learning approaches and novel approaches that aim to more directly harness the key features of spike-based neuromorphic hardware. While conventional feedforward deep neural networks show modest if any benefit on Loihi, more brain-inspired networks using recurrence, precise spike-timing relationships, synaptic plasticity, stochasticity, and sparsity perform certain computation with orders of magnitude lower latency and energy compared to state-of-the-art conventional approaches. These compelling neuromorphic networks solve a diverse range of problems representative of brain-like computation, such as event-based data processing, adaptive control, constrained optimization, sparse feature regression, and graph search.

/pdf/advancing-neuromorphic-computing-with-loihi-a-survey-of-40yqsqbhbl.pdf

Advancing Neuromorphic Computing With Loihi: A Survey of Results and Outlook

Key-value stores such as LevelDB and RocksDB offer excellent write throughput, but suffer high write amplification. The write amplification problem is due to the Log-Structured Merge Trees data structure that underlies these key-value stores. To remedy this problem, this paper presents a novel data structure that is inspired by Skip Lists, termed Fragmented Log-Structured Merge Trees (FLSM). FLSM introduces the notion of guards to organize logs, and avoids rewriting data in the same level. We build PebblesDB, a high-performance key-value store, by modifying HyperLevelDB to use the FLSM data structure. We evaluate PebblesDB using micro-benchmarks and show that for write-intensive workloads, PebblesDB reduces write amplification by 2.4-3x compared to RocksDB, while increasing write throughput by 6.7x. We modify two widely-used NoSQL stores, MongoDB and HyperDex, to use PebblesDB as their underlying storage engine. Evaluating these applications using the YCSB benchmark shows that throughput is increased by 18-105% when using PebblesDB (compared to their default storage engines) while write IO is decreased by 35-55%.

PebblesDB: Building Key-Value Stores using Fragmented Log-Structured Merge Trees

Inspired by physically adaptive, agile, reconfigurable and multifunctional soft-bodied animals and human muscles, soft actuators have been developed for a variety of applications, including soft grippers, artificial muscles, wearables, haptic devices and medical devices. However, the complex performance of biological systems cannot yet be fully replicated in synthetic designs. In this Review, we discuss new materials and structural designs for the engineering of soft actuators with physical intelligence and advanced properties, such as adaptability, multimodal locomotion, self-healing and multi-responsiveness. We examine how performance can be improved and multifunctionality implemented by using programmable soft materials, and highlight important real-world applications of soft actuators. Finally, we discuss the challenges and opportunities for next-generation soft actuators, including physical intelligence, adaptability, manufacturing scalability and reproducibility, extended lifetime and end-of-life strategies. Soft actuators are flexible and compliant and thus perfectly suited to interact with the human body. This Review discusses tethered, untethered and biohybrid soft actuation strategies, highlights promising real-world applications of soft robots and identifies key future challenges, such as implementing physical intelligence and end-of-life strategies.

Soft actuators for real-world applications

From stadium covers to solar sails, we rely on deployability for the design of large-scale structures that can quickly compress to a fraction of their size1–4. Historically, two main strategies have been used to design deployable systems. The first and most frequently used approach involves mechanisms comprising interconnected bar elements, which can synchronously expand and retract5–7, occasionally locking in place through bistable elements8,9. The second strategy makes use of inflatable membranes that morph into target shapes by means of a single pressure input10–12. Neither strategy, however, can be readily used to provide an enclosed domain that is able to lock in place after deployment: the integration of a protective covering in linkage-based constructions is challenging and pneumatic systems require a constant applied pressure to keep their expanded shape13–15. Here we draw inspiration from origami—the Japanese art of paper folding—to design rigid-walled deployable structures that are multistable and inflatable. Guided by geometric analyses and experiments, we create a library of bistable origami shapes that can be deployed through a single fluidic pressure input. We then combine these units to build functional structures at the metre scale, such as arches and emergency shelters, providing a direct route for building large-scale inflatable systems that lock in place after deployment and offer a robust enclosure through their stiff faces. Origami-inspired multistable structures that can be inflated from flat to three dimensions have been designed; a library of foldable shapes is created and then combined to build metre-scale functional structures.

Multistable inflatable origami structures at the metre scale.

Inspired by the embodied intelligence observed in octopus arms, we introduce magnetically controlled origami robotic arms based on Kresling patterns for multimodal deformations, including stretching, folding, omnidirectional bending, and twisting. The highly integrated motion of the robotic arms is attributed to inherent features of the reconfigurable Kresling unit, whose controllable bistable deploying/folding and omnidirectional bending are achieved through precise magnetic actuation. We investigate single- and multiple-unit robotic systems, the latter exhibiting higher biomimetic resemblance to octopus' arms. We start from the single Kresling unit to delineate the working mechanism of the magnetic actuation for deploying/folding and bending. The two-unit Kresling assembly demonstrates the basic integrated motion that combines omnidirectional bending with deploying. The four-unit Kresling assembly constitutes a robotic arm with a larger omnidirectional bending angle and stretchability. With the foundation of the basic integrated motion, scalability of Kresling assemblies is demonstrated through distributed magnetic actuation of double-digit number of units, which enables robotic arms with sophisticated motions, such as continuous stretching and contracting, reconfigurable bending, and multiaxis twisting. Such complex motions allow for functions mimicking octopus arms that grasp and manipulate objects. The Kresling robotic arm with noncontact actuation provides a distinctive mechanism for applications that require synergistic robotic motions for navigation, sensing, and interaction with objects in environments with limited or constrained access. Based on small-scale Kresling robotic arms, miniaturized medical devices, such as tubes and catheters, can be developed in conjunction with endoscopy, intubation, and catheterization procedures using functionalities of object manipulation and motion under remote control.

https://www.pnas.org/content/pnas/118/36/e2110023118.full.pdf

Stretchable origami robotic arm with omnidirectional bending and twisting

Soft robotics has yielded numerous examples of soft grippers that utilize compliance to achieve impressive grasping performances with great simplicity, adaptability, and robustness. Designing soft grippers with substantial grasping strength while remaining compliant and gentle is one of the most important challenges in this field. In this paper, we present a light-weight, vacuum-driven soft robotic gripper made of an origami “magic-ball” and a flexible thin membrane. We also describe the design and fabrication method to rapidly manufacture the gripper with different combinations of low-cost materials for diverse applications. Grasping experiments demonstrate that our gripper can lift a large variety of objects, including delicate foods, heavy bottles, and other miscellaneous items. The grasp force on 3D-printed objects is also characterized through mechanical load tests. The results reveal that our soft gripper can produce significant grasp force on various shapes using negative pneumatic pressure (vacuum). This new gripper holds the potential for many practical applications that require safe, strong, and simple grasping.

/pdf/a-vacuum-driven-origami-magic-ball-soft-gripper-evu2l6fxbo.pdf

A Vacuum-driven Origami “Magic-ball” Soft Gripper

With the advent of large-scale neuromorphic platforms, we seek to better understand the applications of neuromorphic computing to more general-purpose computing domains. Graph analysis problems have grown increasingly relevant in the wake of readily available massive data. We demonstrate that a broad class of combinatorial and graph problems known as dynamic programs enjoy simple and efficient neuromorphic implementations, by developing a general technique to convert dynamic programs to spiking neuromorphic algorithms. Dynamic programs have been studied for over 50 years and have dozens of applications across many fields.

Dynamic Programming with Spiking Neural Computing

The skip list is an elegant dictionary data structure that is commonly deployed in RAM. A skip list with N elements supports searches, inserts, and deletes in O(log N) operations with high probability (w.h.p.) and range queries returning K elements in O(log N + K) operations w.h.p.A seemingly natural way to generalize the skip list to external memory with block size B is to "promote" with probability 1/B, rather than 1/2. However, there are practical and theoretical obstacles to getting the skip list to retain its efficient performance, space bounds, and high-probability guarantees.We give an external-memory skip list that achieves write-optimized bounds. That is, for 0

https://dl.acm.org/doi/pdf/10.1145/3034786.3056117

Write-Optimized Skip Lists

Perhaps the most popular sparse graph storage format is Compressed Sparse Row (CSR). CSR excels at storing graphs compactly with minimal overhead, allowing for fast traversals, lookups, and basic graph computations such as PageRank. Since elements in CSR format are packed together, additions and deletions often require time linear in the size of the graph. We introduce a new dynamic sparse graph representation called Packed Compressed Sparse Row (PCSR), based on an array-based dynamic data structure called the Packed Memory Array. PCSR is similar to CSR, but leaves spaces between elements, allowing for asymptotically faster insertions and deletions in exchange for a constant factor slowdown in traversals and a constant factor increase in space overhead. Our contributions are twofold. We describe PCSR and review the theoretical guarantees for update, insert, and search for PCSR. We also implemented PCSR as well as other basic graph storage formats and report our findings on a variety of benchmarks. PCSR supports inserts orders of magnitude faster than CSR and is only a factor of two slower on graph traversals. Our results suggest that PCSR is a lightweight dynamic graph representation that supports fast inserts and competitive searches.

Packed Compressed Sparse Row: A Dynamic Graph Representation

Various applications model problems as streaming graphs, which need to quickly apply a stream of updates and run algorithms on the updated graph. Furthermore, many dynamic real-world graphs, such as social networks, follow a skewed distribution of vertex degrees, where there are a few high-degree vertices and many low-degree vertices. Existing static graph-processing systems optimized for graph skewness achieve high performance and low space usage by preprocessing a cache-efficient graph partitioning based on vertex degree. In the streaming setting, the whole graph is not available upfront, however, so finding an optimal partitioning is not feasible in the presence of updates. As a result, existing streaming graph-processing systems take a "one-size-fits-all" approach, leaving performance on the table. We present Terrace, a system for streaming graphs that uses a hierarchical data structure design to store a vertex's neighbors in different data structures depending on the degree of the vertex. This multi-level structure enables Terrace to dynamically partition vertices based on their degrees and adapt to skewness in the underlying graph. Our experiments show that Terrace supports faster batch insertions for batch sizes up to 1M when compared to Aspen, a state-of-the-art graph streaming system. On graph query algorithms, Terrace is between 1.7X--2.6X faster than Aspen and between 0.5X--1.3X as fast as Ligra, a state-of-the-art static graph-processing system.

https://dl.acm.org/doi/pdf/10.1145/3448016.3457313

Helen Xu

Papers

A Vacuum-driven Origami “Magic-ball” Soft Gripper

Dynamic Programming with Spiking Neural Computing

Write-Optimized Skip Lists

Packed Compressed Sparse Row: A Dynamic Graph Representation

Terrace: A Hierarchical Graph Container for Skewed Dynamic Graphs